What is the format of the query.expected files?

[gagliardetto]

I'm writing a parser for the .expected files (I will then use the data to visually represent the query results).

• Is there some documentation on the format of the .expected files?

• Anything in particular I'd need to know when parsing them?

• Also, is the .expected file the best source for getting the edges/nodes data?

Thanks!

[smowton]

To quote @aibaars from Slack:

It's just plain text. Usually a table-like format like | value1 | value2 | ... | If you use named query predicates (query predicate fooBar) then each "table" is prefixed with a "title" like

fooBar
| value1 | value2 | ... |

The expected output may also contain compiler output such as errors and warnings in addition to the table output.

The above is the case for most tests. However, some query @kinds have different output. For example @kind graph or @kind tree produce some textual graph representation instead of a table.

(Paraphrasing now), for most purposes it is better to run a query with a more firmly-defined format like CSV, or for graph queries DOT or DGML, than to try to parse the .expected format that may change from version to version.

[gagliardetto]

Thanks!